Sorry boys, Grendel is right
I'm one of the people at WotC who contributed to the design of Tooth & Claw, and who worked on developing ECLs for the creatures for the Monster Manual. To answer some of your questions:
• No, Zerovid, we didn't have a stupid set formula to use. We considered each monster individually and in comparison with other monsters. Then I did some playtesting, from 1-on-1 fights to party composition tests.
• Grendel is dead on with our philosophy. It's already way cool to be able to play a monster (ogre, doppleganger, or lizardfolk, to mention those discussed here), and so it shouldn't also be more powerful than the human PCs in a party. Who would want to be the plain old human if the ogre is not only more interesting, which he'd just about have to be, but also just better?
• Let's talk about the lizardfolk. Some of you noted that the lizardfolk has 2 Hit Dice, and that's important. Your ECL has to *at least* equal your Hit Dice, I hope everyone realizes. And if it does equal your Hit Dice, you should get about, well, nothing else. Now, let's look at what you'd get for playing a lizardfolk. Stats: +2 STR, CON +2, INT –2. That's a very good deal for a fighter. Then there's the +5 natural armor. Equal to 50,000 GP in magic items, and it means that this PC is going to have much higher AC, especially at lower levels when no one can afford a big amulet of natural armor (the lizardfolk fighter can still wear full plate or anything else that a human fighter can). Wouldn't you like to be hit something like 25% less often? Well, I would. Then add the skill bonuses,which aren't terribly important, but they're some nice gravy.
So anyone who wants to give them ECL 2 needs to think again. That's essentially saying that you can make take the half-orc and change his –2 Charisma penalty to a +2 Constitution bonus, add 5 points of natural armor, and that you'd keep him ECL 0. Uh, no.
Now, those of you who thought ECL 4 is a little too high, let's talk. Maybe you think that the bonus to ability scores and natural is worth about a level, not two. So you'd make the lizardfolk ECL 3. I could see merit in that argument; and in fact I might be tempted to agree. I personally think the lizardfolk with the abilities I've listed off is a bit too good at ECL 3, and perhaps a bit too bad at ECL 4. That said, I'd lean toward ECL 4 because though I think playing monsters is cool, and I love doing it and seeing my players do it, I never want someone who plays one of the standard races from the PHB to feel so overpowered, to feel dumb for making that choice.
But we forgot something! The lizardfolk, on top of everything else, has a claw/claw/bite routine. That means that, before adding anything else, this character is going to be able to make 3 attacks per round. Sure, those attacks don't do that much damage really individually (though they will on a PC lizardfolk, which can easily have 20 STR), but a little knowledge of the D&D game will teach you that it's better to have an extra attack with one or two less points of damage. Moreover, think about the lizardfolk with just 1 rogue level. Get this guy a flank and he's swinging three times per round, with three possible sneak attacks. That's the sauce, people.
When you consider a monster and its ECL, be careful to consider every line of the statistic block, every ability score adjustment, and every special ability. It's a complex thing, part art and part science. You have to consider what those abilities will mean in a party, where they may have been less relevant to a monster encountered in a room and defeated and never seen again. Lots of things go into the process, and it took a team here hundreds of total man-hours to work up ECLs you're seeing.
That's not to say we can't make mistakes. It is true that some of the ECLs may be too high, and some too low. We continue to playtest and discuss them here in preparation for the book Tooth & Claw.
David Eckelberry
R&D, Wizards of the Coast